Monitoring and strengthening interventions on the

Structural Analysis of Historic Construction – D’Ayala & Fodde (eds)
© 2008 Taylor & Francis Group, London, ISBN 978-0-415-46872-5
Monitoring and strengthening interventions on the stone tomb of
Cansignorio della Scala, Verona, Italy
G. Gaudini
Soprintendenza per i Beni Architettonici e per il Paesaggio di Verona, Vicenza e Rovigo, Verona, Italy
C. Modena & F. Casarin
Department of Structural and Transportation Engineering, University of Padova, Italy
C. Bettio & F. Lucchin
SM Engineering, Padova, Italy
ABSTRACT: The “Arche Scaligere” is the gothic monumental funerary complex of the illustrious Veronese
“Scaligeri” family. Between these monuments, the “Arca” (stone tomb) of Cansignorio della Scala is the most
sumptuously decorated. In 2005 the “Soprintendenza per i Beni Architettonici e per il Paesaggio di Verona,
Vicenza e Rovigo” (the local administrative body in charge of the conservation of the landscape, monuments and
historical structures), with the aim of studying the stone tomb of Cansignorio and in the framework of a wider
research, gave the task of designing and installing a structural monitoring system, besides the implementation of
a detailed FE numerical model, to the University of Padova. In parallel with these studies, a light strengthening
intervention was carried out in order to stabilize some critic points of the structure and to intervene in deteriorated
parts or elements, to integrate the original materials. The paper describes the different activities carried out.
1
1.1
FOREWORD
Historical notes and description
The “Scaligeri” or “della Scala” family was a dynasty
that ruled Verona, in Italy, for over a century, from
1262 to 1387. Cansignorio della Scala (1340–1375)
managed the city in a relatively peaceful period, and
adorned Verona in a way to make it call “marmorina”
(marbled) for the abundant use of ancient marbles and
roman statues.
The stone tomb of Cansignorio della Scala was built
between 1374 and 1376, by will of the same Cansignorio, when he was still alive. The tomb was erected close
by the St. Maria Antica church, where the tombs of
Cangrande and Mastino the 2nd (his grandfather and
father respectively) were already built by local workers. Differing from his ancestors, Cansignorio desired
a monumental tomb, where the architectural aspect
was more important than the decorative. The work was
then commissioned to Bonino da Campione, a famous
master of gothic sculpture. The monument, based on a
hexagonal plan, is adorned with sculptures and spired
tabernacles, with the overhanging equestrian statue of
Cansignorio. The tomb is surrounded by an hexagonal
wrought iron fence, at whose corners rise six pillars
Figure 1. The stone tomb of Cansignorio (on the right) and
Mastino the 2nd (rear left), near the St. Maria Antica church.
403
1.2
Figure 2. The upper part of the Cansignorio stone tomb:
below the gablets with the statues of the virtues and the
tabernacles with angels; above the equestrian statue.
sustaining gothic tabernacles, containing statues of the
saint-warriors (St. George, St. Martin, St. Quirinus,
St. Sigismund, St. Valentine and St. Louis, king of
France). The tomb starts with six columns sustaining a
red marble slab on which finds place the white marble
sarcophagus, sustained by eight pillars and decorated
with bas-reliefs representing Gospel scenes. The cover
of the sarcophagus hosts a lying statue of Cansignorio,
watched over by angels.
At the second level, six further spiral columns
sustain the canopy with polylobed arches. Above
these finds place a cornice sustaining six gablets
with allegorical figures representing the virtues. At
the corners are positioned six further tabernacles
with statues of angels. The roof, corresponding to
an hexagonal pyramid made of white marble, finally
supports the massive equestrian statue of Cansignorio
(Fig. 2).
The stones used for the erection of the tomb are
the “Candoglia” white marble, the same employed
in the Milan’s cathedral, and the “Rosso di Verona”
(Verona’s red marble), besides the Pietra Gallina (a soft
limestone from Vicenza). The inner part of the roof
(above the crossed vault and behind the stone facing of the canopy) is composed by solid brickwork
masonry.
Past restoration works
Throughout the centuries, several repair interventions
were necessary to preserve the delicate structure of
the stone tomb, such as those carried out in the XVII,
XIX and XX centuries. In 1676 the Verona municipality adopted a resolution to execute restoration works on
the tomb, comporting strengthening interventions and
substitutions on the upper part of the monument, without however intervening on the supporting elements.
Between 1827 and 1829 other restoration works were
carried out, raising arguments on the type of marble
to be used in substitutions of the deteriorated parts.
Between 1838 and 1844 the fence was restored, and
on the 24th of July 1840 a portion of the southern
gablet fell down, being subsequently restored (1846)
and lodged back in the original position. Substitutions comported the use of Candoglia marble elements,
secured with iron clamps fixed with melted lead. The
sealing of the cracks was performed with filler. Main
interventions carried out were: the reconstruction of
the spires of some tabernacles; the positioning of steel
reinforcing elements on two columns of a tabernacle;
the complete reconstruction of a column and capital of
a tabernacle, and of some gablets between the spires;
the reconstruction of the tail and the left rear leg of the
horse in the equestrian statue; the substitution of the
copper tie beams of the tabernacles with saint-warriors
with new ones in iron; the sealing of the vault’s groins.
Other interventions, similar to those executed at the
half of the XIX century, were carried out between 1910
and 1914. The monument was then protected against
bombing during the two world wars. In 1919, after the
removal of the shields, some light restorations were
carried out. Then, during the positioning of the shields
of the 2nd world war, an analysis of the conditions
of the tombs was carried out, with successive light
restoration works.
2 THE INVESTIGATION ACTIVITIES
2.1
Dynamic identification
Between different Non Destructive techniques that
may be profitably used for the achievement of an
advanced knowledge of the structural layout of a historic masonry building, dynamic identification proved
to be a very effective tool (Modena et al., 2001; Gentile
et al., 2004; Ramos et al., 2006), being actually the
only method able to experimentally define parameters
related to the global structural behavior. Prior to the
installation of a Structural Health Monitoring (SHM)
System, a dynamic investigation campaign took place
in August 2006. Tests were aimed at the definition of
the optimal SHM system sensors’ positioning, and at
the characterization of the dynamic properties of the
monument for FE modelling calibration purposes.
404
Figure 3. Identified mode shapes: (a) 1st bending N-S, 3.19 Hz; (b) 1st bending E-W, 3.24 Hz; (c) 1st torsion, 5.88 Hz;
(d) 2nd bending N-S, 12.55 Hz; (e) 2nd bending E-W, 12.88 Hz; (f) 2nd torsion, 19.42 Hz; (g) FDD method, average of the
normalized singular values of spectral density matrices of all test setups.
Following the mode shapes emerged from the FE
numerical model, sensors were placed at the first level
(in the stone slab where the sarcophagus stands), at the
second level (on the cornice above the pointed arches)
and at the top of the monument (at the foot of the
equestrian statue). A total of six sensors was employed,
considering three test setups for a total of 6 acquisition
points, recording the acceleration in orthogonal (and
parallel to the ground) directions.
The acquisition system was composed by a compact unit provided with 24-bit digital acquisition cards,
connected to piezoelectric mono axial acceleration
transducers. Once fixed the transducers to the structure
in the selected positions, tests consisted in acquiring
data over a predetermined period, at a determinate
sample rate. Each test setup consisted in recording
the signal two times (65,536 points each) whit a sampling frequency of 100 SPS (samples per second), with
an overall setup signal recording duration of 21 51 .
For the identification of the modal parameters (natural
frequencies and corresponding mode shapes), output
only identification techniques were used (Operational
Modal Analysis). In particular, the recorded ambient vibrations were related to the wind excitation and
urban traffic.
The modal parameter extraction method selected
was the FDD – Frequency Domain Decomposition –
technique (Brincker et al. 2000) which estimates the
modes, with the assumption that the excitation is reasonably random in time and in the physical space of
the structure, using a Singular Value Decomposition
(SVD) of each of the spectral density matrices. The
data series acquired at 100 SPS were processed by a
decimation of 2 (Nyquist frequency of 25 Hz), with
segment length of 2048 points and 66.67% window
overlap. Several peaks related to structural frequencies were detected in the frequency domain and the
corresponding mode shapes defined (Fig. 3).
2.2
Monitoring
The Structural Health Monitoring System (installed
in December 2006) is aimed at the control of static
405
corresponding to 4 daily readings. Dynamic data are
collected both at fixed time intervals (each 48 hours,
approximately 22’ of recording at a sample rate of 100
Hz) and on a trigger basis (shorter records, signals
are recorded when the vibration exceeds a predefined
threshold).
No meaningful variations in terms of displacements
were reported up to November 2007 (Fig. 5a). Variations remain limited and related to the environmental
parameters, presenting maximum differences (corresponding to crack mouth opening) of about 1/10th of
millimeter. No seismic events were recorded in the
monitored period. Limited shifts (max 4%) were noted
in all of the identified frequencies (see also Ramos
et al., 2007), possibly related to environmental parameters, as reported in Figures 5b and c (slight decrease
with the relative humidity, seasonal variations).
Figure 4. Positioning of the acceleration (left) and displacement sensors (right).
and dynamic parameters related to the structural functioning of the monument. The system is composed
by an acquisition unit connected to six piezoelectric accelerometers, two potentiometric displacement
transducers and a temperature and relative humidity
sensor. The central unit, located at the base of the
tomb, is provided with a WiFi router for remote data
transmission.
The monitoring strategy is conceived both to collect
data at predetermined time-intervals (periodic monitoring, i.e. cracks opening, changes in the dynamic
response) and to automatically start to save data in case
of significant external events (such as seismic events).
Such controls will permit to appreciate possible variations in the assessed structural functioning with the
passing of time and to have a record of the dynamic
behavior of the stone tomb during severe events.
The acceleration transducers are placed in suitable
positions in relation to the mode shapes of the structure, as shown by the numerical modeling/dynamic
identification (Fig. 4, left). Four sensors are placed
on two levels for the evaluation of the vibration in
the NS and EW direction (bending modes) and in the
horizontal planes (torsion modes).
A couple of reference sensors is fixed at the base
for the record of the ground acceleration both in operational conditions (i.e. evaluation of the traffic induced
vibrations) and during seismic events. A temperature/relative humidity sensor is fixed at the intrados
of the marble slab (first level). The displacement
transducers are positioned across significant cracks
(Fig. 4, right, see also Fig. 14). The temperature, relative humidity and displacement of the selected points
(crack mouth opening) are recorded each 6 hours,
3
3.1
STRUCTURAL MODELS
Introduction
A detailed FE numerical model, based on a laser scanner geometrical survey of the monument previously
carried out, was implemented in order to evaluate the
static and dynamic behaviour of the monument. The
evaluation of the initial results of the numerical model
(linear static and natural frequency analyses) assisted
the design phase of the strengthening intervention and
indicated the most suitable places for the sensors’positioning (dynamic identification and monitoring). The
first model was calibrated on the basis of the results
of the experimental activities, in order to be subsequently used to simulate the response of the monument
to different external actions.
3.2 The FE model
As a first step, linear elastic constitutive laws were
assigned to all materials in order to define the static
load pattern (self weight) and the dynamic properties
of the monument. The model is composed by approximately 49,000 brick elements and 53,600 nodes. Finite
elements’ sides are comprised between 0.10–0.15 m.
The mesh is more refined in the slender elements
(columns) and in the junctions, rougher elsewhere. The
decorative elements and statues were modelled as the
structural parts: only areas too small to be considered
significant were neglected (Fig. 6).
The linear static analysis (self weight) indicates
that compressive stresses reach their maximum values
in the columns, where stresses of about 1.0–1.5 MPa
(lower order) are found. In small areas of the upper
order of columns compressive stress peaks of 2.0 MPa
are noted. Tensile stresses generally present very low
values or close to zero. However, non negligible tensile
406
Figure 5. Monitoring results: (a) displacement transducers PZ1/PZ2 and environmental parameters, recorded data plotted
vs. time; dynamic parameters, identified frequencies: (b) first two bending frequencies vs. time and (c) vs. relative humidity.
Table 1.
Experimental vs. numerical frequencies.
Frequency (Hz)
Figure 6. (a) Rendered view of the FE model, East side;
(b) corresponding mesh. The positive Y axis corresponds to
the North direction.
Mode
Description
Exp.
FE model
diff. %
1
2
3
4
5
6
1st bending N-S
1st bending E-W
1st torsion
2nd bending N-S
2nd bending E-W
2nd torsion
3.19
3.24
5.88
12.55
12.88
19.42
3.25
3.26
5.85
12.90
13.30
19.38
1.88
0.71
0.48
2.79
3.26
0.21
stresses localize at the crown of the pointed arches and
on the above cornices.
The natural frequency analysis indicates the elastic
dynamic characteristics of the structure (natural frequencies and mode shapes). In the calibrated FE model
8 sets of materials are considered, theYoung’s modulus
ranging from 40,000 (solid stone) to 4000 MPa (brickwork masonry), the corresponding densities from 2700
to 1900 kg/m3 . Six principal modes emerged from
the analysis (Fig. 7). A close match between the frequencies/mode shapes emerged from the calibrated
numerical model and those emerged from the dynamic
investigation was found (Table 1).
407
Figure 8. Strengthening of the stone tomb of Cansignorio
della Scala, detail of the interventions.
Figure 7. 1 to 6, first six mode shapes of the numerical
model.
4 THE STRENGTHENING INTERVENTION
4.1
Introduction
supports (horse’s hooves) of the equestrian statue of
Cansignorio by means of CFRP strips; (2) strengthening a cracked capital with hoopings in high resistance
stainless steel cable.
4.2 Interventions description
In parallel with the studies previously reported, and
beneficiating from their outcomes, a light strengthening intervention was carried out. New structural
elements were introduced as precautionary measures,
e.g. by providing redundant confining systems, to collaborate with existing deteriorated elements and acting
in case of sudden structural deficiency of the original
material.
In general the stone tomb does not present indications of worrying structural problems. Interventions
mainly consisted in hooping the monument at different levels. With reference to Figure 8, interventions
included: A) hooping the base of the canopy with
a stainless steel cable; B) hooping the capitals with
a couple of stainless steel cables; C) repair of the
junctions of existing tie beams; D) hooping the tabernacles with high resistance stainless steel cable. Local
interventions consisted in: (1) binding the damaged
Widespread cracks were noted in the cornice above the
pointed arches, and in the same arches close by their
keystones (Fig. 9a, b).
Parts of the cornices tended to separate, being this
however not a recent damaging process, since iron
clamps of previous interventions were found. A hooping device (hooping A) consisting in a stainless steel
7 mm diameter cable, connecting 6 corner steel plates
and tensioned by turnbuckles, was positioned above
the cornices. The size of the steel elements was minimized in order not to be invasive with respect to the
monument (Fig. 9c, d).
The existing iron tie beams, spanning between the
capitals of the upper order columns and locally damaged by oxidation (Fig. 10a), were complemented with
a couple of 3 mm diameter stainless steel cables (hooping B). In fact, a strengthening intervention on the
original tie beams was not feasible without heavily
408
Figure 9. Hooping A, details.
Figure 11. Intervention C (a, b) and hooping D (c, d),
details.
Figure 10. Hooping B, details.
intervening on the columns capitals. Cables were fixed
to the capitals through stainless steel bushes, moulded
following the shape of the capitals (Fig. 10c). An even
contact between steel and stone was provided by means
of lead sheets.
The iron tie beams connecting the tabernacles of
the fence to the spiral columns, likely positioned during the XIX c. interventions, manifested marked decay
at the connection with the original copper tie beams
anchored to the stone. The tie beam strengthening
intervention (C) aimed at the restoration of the original elements avoiding the onset of a new oxidation
process on the copper anchoring elements. Two titanium studs were placed to join the existing iron and
copper tie beams (Fig. 11a, b).
The deterioration or lack of the original iron
tie beams in the same tabernacles required the
Figure 12. Strengthening of the horse’s hooves: a) copper
bandage; b) deteriorated stone conditions; c) application of
CFRP strips; d) final appearance.
introduction of new elements (high resistance stainless steel 1.6 mm diameter cables – hooping D) to
restore the original layout (Fig. 11c, d). To minimize
the dimensions of the clamps, the fixing methodology
was tested (tensile strength of cable and connection)
by means of laboratory experimental activities.
The removal of the copper “bandage” provided to
the equestrian statue of Cansignorio during past interventions for the strengthening of the supports of the
statue, highlighted the presence of material decay, with
severe cracks and voids (Fig. 12a, b). Damaged supports were strengthened by the application of high
resistance CFRP strips, subsequently covered with a
plaster facing (Fig. 12c, d).
409
Figure 13. Strengthening of one of the stone capitals of the
spiral columns: (a) damage induced by the tie beam oxidation;
(b) intervention layout.
Figure 14. Above: a almost full-scale picture of the used displacement transducers; below: displacement transducer PZ2
prior (left) and after (right) the restoration intervention.
The expansion of the iron tie beams due to material oxidation caused the cracking of the capitals of the
upper order columns, in some cases with severe effects
(Fig. 13a). The strengthening intervention required
the sealing of the crack and the positioning of hoopings on the capital, on 3 levels (1.6 mm diameter high
resistance stainless steel cable). Purposely shaped titanium elements (Fig. 13b) were employed to allow the
clamping of the cable.
5
CONCLUSIONS
Extensive studies were carried out on the Arca of
Cansignorio della Scala, in Verona, in parallel with a
light structural strengthening intervention and stone
restoration activities. The research involved several
aspects, some of them not reported in the paper (e.g.
the laser scanner survey or stone characterization analyses), aimed at the achievement of a complete picture
of the monument, for conservation purposes.
The investigation activities carried out (dynamic
identification) proposed experimental evidences for
the calibration of behavioral models of the monument.
Finite Elements models of the Arca were implemented on the basis of the laser scanner survey of
the monument. Models were used to predict the static
and dynamic behavior of the building and were successively tuned on the basis of the experimental activities.
In a successive stage the models will be considered
(with material non linear properties) to assess the
response of the monument to seismic events.
The installation of a static and dynamic Structural
Health Monitoring System gives the possibility to continuously evaluate the conditions of the monument
by recording significant indicators (environmental
parameters, dynamic response, cracks opening). The
systems also allows to check the dynamic response of
the structure to traffic or seismic events. An important aspect considered in the setup of the monitoring
system was the reduced impact on the monument,
given also the continuous attendance of tourists: sensors were minimized and “camouflaged” as much as
possible (Fig. 14).
The design of the intervention was based on a almost
complete removability of the new structural elements,
positioned to compensate the lacks of the original
material. Interventions, except the consolidation of
heavily damaged parts (horse’s hooves), are based on
a mechanical assembly of metallic elements, avoiding chemical connections with the original material.
New materials and structural elements were chosen in
order to maximally reduce their dimension (i.e. high
resistance stainless steel cables).
ACKNOWLEDGEMENTS
The authors would like to thanks student Manuel
Marotto for his valid collaboration in the execution
of the research activities.
REFERENCES
Bertolini V. Da Nantes a Brou e a Ginevra: influssi possibili o
irrefutabili delle Arche Scaligere. 1993, Verona (in Italian)
Brincker, R., Zhang, L., Andersen P. Modal Identification from Ambient Responses using Frequency Domain
Decomposition, Proc. 18th International Modal Analysis
Conference, 2000, San Antonio, TX
De Maffei F. Le arche scaligere. 1955, Verona (in Italian)
Gentile C., Saisi A. Dynamic-based F.E. Model Updating to
Evaluate Damage in Masonry Towers. Proc. 4th Int. Seminar on Structural Analysis of Historical Constructions,
2004, Padova, Italy
Modena C., Franchetti P., Zonta D., Menga R., Pizzigalli E.,
Ravasio F., Muti M., Meloni R. and Bordone G. Static and
Dynamic Analyses of Maniace Castle in Siracusa-Sicily.
Proc. 3rd Int. Seminar on Structural Analysis of Historical
Constructions, 2001, Guimarães, Portugal
410
Napione E. Le Arche Scaligere di Verona. Storia di un sepolcro dinastico, 2000–2003 tesi di dottorato di ricerca in
Storia dell’arte, Università degli studi di Udine (in Italian)
Napione E., Varanini G.M. «Gaspar recultor» e l’Arca di Cansignorio della Scala, in Verona illustrata n.18. Museo di
Castelvecchio, 2005, Verona (in Italian)
Ramos L., Casarin F., Algeri C., Lourenço P.B., Modena C.
Investigation techniques carried out on the Qutb Minar,
New Delhi, India. Proc. 5th Int. Seminar on Structural
Analysis of Historical Constructions, 2006, New Delhi,
India
Ramos L., Marques L., Lourenço P., De Roeck G.,
Campos-Costa A., Roque J. Monitoring historical
masonry structures with operational modal analysis: two
case studies. 2nd International Operational Modal Analysis Conference, 2007, Copenhagen, Denmark
411